Build your first personal financial system

You have the ideas. You don't yet have the thing.

Six chapters in, you know what a ledger is and what it is for. You know money comes in, moves, and goes out. You know what your accounts really are, what counts as income, and what a budget you might actually keep looks like.

And you still have an empty app. That gap is where most people stop, and they stop in one of two ways:

Both failures have the same cause: nobody told you how small this is allowed to be. So here is the whole of it — the smallest setup that actually works, in the order the pieces depend on each other. About half an hour, and you can stop halfway and finish next weekend without breaking anything.

Half an hour, and the order matters

This is the part worth screenshotting. Five steps, and they are in this order for a reason: each one exists because the next one needs it. Categories have nothing to attach to until there are accounts. A budget has nothing to plan into until there are categories. And no number anywhere means anything until you have said which currency it is in.

Honest timings, for someone who has not done this before: about half an hour if you do it in one sitting. Steps 4 and 5 can wait until next weekend — with the first three done, the app already works. It will record what you spend and tell you where you stand. Categories and a budget make it tell you more.

You will notice what is not on that list: importing statements, connecting banks, tags, recurring bills, a second country. All of those are real, and all of them are better done later. §8 says exactly when.

Step 1 · Household: who this money is for

The least obvious of the five, so let's define it before it confuses anyone. A household is the unit the money belongs to — not the person typing.

If you live alone, you are a household of one, and it takes two minutes and no thought. The concept still earns its place, because of what it makes possible the moment there is more than one of you: a couple can both see the same rent payment, the same grocery budget, the same set of accounts, without either of them maintaining a private second copy of it. The rent leaves the household once, and it is recorded once.

The practical consequence is worth holding on to, because it explains a lot of the app's behaviour later: everything you create from here belongs to the household, not to your login. Accounts, categories, budgets, receipts. Your login is how you get in. The household is what you get into.

Step 2 · Your ledger, and the currency you think in

A ledger is one country's financial life. One ledger per country, and its home currency is chosen once, at creation, and never changes afterwards.

People agonise over that choice for longer than it deserves, so here is a decision rule instead of a lecture:

Pick the currency your salary lands in and your rent leaves in.Not the one you grew up with. Not the one your savings happen to sit in. The one your month is denominated in.

That rule settles nearly every case. If you live and work in Hong Kong, it is HKD, even if you are Korean and still think in won, and even if half your savings are in dollars. If you have just moved, it is the new country — and your old country becomes a second ledger later, when you get round to it, which is a different chapter and not today's problem.

Two small things you do not have to worry about. Spending in another currency is fine and normal — the receipt keeps its own figure and the ledger shows you yours. And a second country later does not mean redoing any of this; it means a second ledger sitting beside this one. Why Aurora keeps them side by side rather than adding them together is the argument of the next chapter.

Step 3 · Accounts: the ones money actually moves through

This is where the Sunday over-build starts, so this section is going to be blunt with you.

Add the account your salary arrives in. Add the card you spend on. Add cash if you actually use it. That is usually three, and three is a complete answer.

Chapter 4 already explained what each account type is. This step only answers a much narrower question: which ones today?

Each account you add wants two things: a starting balance and the date you are starting from. Use today, and use whatever the bank app says right now. And let's be plain about what that means, because it is the single biggest relief in this chapter:

You are not reconstructing history. You are drawing a line and beginning there.Everything before that line stays where it is — in your bank's records, where it has always been.

The one detail worth typing carefully is the last four or five digits of each card. It takes five seconds and it is what lets Aurora tell later on which card a receipt was paid with — which turns out to matter a great deal in Chapter 9.

Step 4 · Categories: eight to fifteen you could name from memory

Chapter 6 gave you the number: eight to fifteen lines you could recite without looking. Enough to be specific about where the money went, few enough that you will still be keeping them in March.

Here is the part Chapter 6 did not tell you, and it is the good part. You mostly do not have to do this. When you created the ledger, Aurora seeded a starter set appropriate to the country you picked. Your job is not to author a list. It is to edit one:

Only one structural mistake is available here, so it is worth one line: you budget the leaves, never the parents. Groceries and Dining Out take numbers; the Daily Living row above them adds itself up and is not something you can type into. The app enforces this rather than trusting you to remember it — which is why a parent row has no amount field at all.

And do not try to get this right. You cannot, yet. A category you never use is deleted in two seconds three weeks from now, and one you are missing announces itself the first time something does not fit.

Step 5 · A budget, in its smallest working form

Last step, and it is deliberately under-scoped. Chapter 6 gave you the whole method — save first, size from your own history, spread it across the year. Today you are doing the two-minute version of it:

Which leaves the budget visibly unfinished, with money still unallocated. That is fine. That is the intended state:

The reason is simple enough to state in one line: a plan you finished is worth more than a plan you refined. A rough budget that exists starts telling you things this week. A perfect one you are still tuning tells you nothing at all.

And there is a proper escape hatch, which is why guessing today costs so little. After a few weeks of real spending, Aurora can size the remaining lines from your own history — a year-to-date average per category, drawn from what you actually did. That number will be better than anything you could invent this afternoon, so let it do the work.

What you deliberately don't do today

Everything on this list is a thing a diligent person attempts in hour one and drowns in. None of it is wrong. All of it is better later — and for most of it, later means when the app tells you it is time.

Read that column again if you skimmed it. The reason a setup like this survives is not that it was thorough. It is that it was small enough to finish, and nothing in it was a chore.

The first two weeks, and what fills itself in

Here is the thing nobody tells you about setting up a ledger: it is not finished when you close the form. It finishes itself while you use it.

Almost everything interesting about the app switches on by itself, once there is enough for it to talk about. You do not configure any of this:

All of which rests on one small habit, and only one: get what you spend into the app. There are three ways in and they take the same few seconds — snap the receipt on your way out of the shop, forward the confirmation email, or upload the statement when it arrives. Any of them, in any order, and no order is the correct one.

Miss a few. Everybody does. A ledger with most of your spending in it is enormously more useful than no ledger, and the app is built to be caught up with rather than kept up with.

What “done” looks like

A month in — with the second one already running — the payoff is not a completed spreadsheet. It is that the app answers questions you never set it up to answer:

Five decisions, made once, doing work every day afterwards. That is the whole trade this chapter is offering.

Your journey so far